Model of wind-generated ambient noise in stratified shallow water

Chinese Journal of Oceanology and LimnologyVol. 23 No.2, P. 144-151, 2005Model of wind-generated ambient noise in stratified shallow water*LIN Jianheng (林建恒), GAO Tianfu (高天赋)*(Qingdao Laboratory of Istite of Acoustis, the Chinese Academy ofSciences, Qingdao 266023, China)(tInstitute ofAcoustics, the Chinese Academy of Sciences, Bejing 100080, China)Received July 8, 2003; revision accepted Dec. 28, 2004AbstractIn order to build a rapid ocean ambient noise model adapted for a stratified shallow water, a hybridmodel of normal mode method (for far field) and ray method (for near field) is suggested which combines theadvantages of both methods. Since the near field of wind-generated noise is not sensitive to the sound speed pro-file, the sound speed profile is regarded as a constant; which makes the model rapid and accurate. The simulationresults are in agreement with those of the wave model.Key words: wind-generated ambient noise, ray, normal mode, hybrid model1 INTRODUCTIONThe outline of the noise model and correlationfunction (power spectrum is included) equations areOcean ambient noise is a main interference withpresented in Section 1; Section 2 gives the simpli-respect to the active and passive sonar, and wind-gen-fied equations of the correlation function. Verticalerated ambient noise is major noise source in maindirectionality model is presented in Section 3. Insonar firequency band from several hundred hertz to 20Section 4, a comparison study of noise intensity andthousand hertz. Developing a rapid and accurate nu-vertical directionality predicted by this model withmerical model to satisfy the need of sonar design andthose of the wave model is shown. Section 5 con-use is an important goal in recent years.tains the summary and conclusions.Since Cron and Sherman (1962) built up a deepsea wind-generated ambient noise model, many re-2 WIND-GENERATED AMBIENT NOISEsearches have been done (Talham, 1964; Liggett andMODEL AND CORRELATION FUNC-Jacobsen, 1965; Buckingham, 1980; Kuperman andTION OF NOISE FIELDIngenito, 1980; 1988; 1996; Hamson,1985; Hamson,1983, 1985, 1997; Harrison, 1997; Yang and Yoo,As shown in Fig.l, we assumed a homogene-1997; Finette and Heitmeyer, 1996) with emphasison shallow water situation. Among these models,the wave model (Kuperman and Ingenito, 1980;Surface aaHamson, 1983, 1985, 1997; Yang and Yoo, 1997)and the ray model (Harrison, 1997) were the mostSurface plane(r;z)applicable for sonar design. In terms of propagation,the wave model is an ideal model, but it takes too(nz)much time for continuous spectrum calculation fornear sources; accurate but slow. The ray model is(z)rapid, but not so accurate on the other hand, as itapplies much approximation for far sources.PInc(z)A hybrid model of normal mode method (for farBottom中国煤化工field) and ray method (for near field) is suggested in thispaper, and for the near field is not sensitive to soundMYHC N M H Ged urfaice noise poblemspeed profile (SSP), and is decided mainly by bottomreflection loss. In our model, the non-uniform SSP* This work was supprted by Naval Weapon Department under contractwas replaced with uniform SSP.No.41303090207. and by Science and Technology Ministry under con-tract No. 2001AA631080 and No. 2003AA604010.No.2LIN et al: Model of wind-generated ambient noise in sratified shallow water145ous statistical distribution of wind-generated noiseG-=2"(")",(2)H"(k,n)+CSC(9)sources S(r',z) on infinite plane z=z' (Kupermanand Ingenito, 1980), the correlation function of sur-where ψ(2) is the normal mode eigenfunc-face noise sources is in the form of:tion; H"(k,r) is the Hankel function; CSC repre-=q'N(r"-r*",z")(1)sents the contribution of continuous spectrum. Thesecond term ofEq. (7) iswhere q^ is surface source strength, andN(r'-r")=4m-8(r'-r")(2)k2(z7)'"k(z)!(10)The noise field for a receiver at (r,z) can bewritten as the total contributions of all individualwhere G is the ray representation of Green's func-sourcestion; its integral expression for a constant soundspeed profile is4(r,z)= jS(',")G(,z;r',z)d'r'(3)where ψ is the displacement potential function of8元π。noise field; G is Green's function, which satisfiesthe Helmholtz equationz'40°)the near field and normal mode method for the farthat are dominated by overhead noise unlike that infield. The noise level scale in Fig.8b and Fig.9 is ahard bottom case. Negative-high-angle beams arerelative level. It is shown that our model and Yang'sassociated with bottom bounced rays, and no directagree well with each other. In computation time,and surface-reflected rays. When grazing angle >34,Yang中国煤化工. group of parame-the noise intensity for soft bottom is weaker thanters 0e our model takesthat for hard bottom.only:YHCNMHGparametersonaFig.8 compares the noise vertical directionalitypersonal computer. The latter is more appropriatefor the constant SSP and that for the summer SSP atfor practical use with high speed and accuracy.500 Hz for soft bottom. (a) is the result of our model,For summer SSP, the interaction between ray.150CHIN. J. OCEANOL. LIMNOL., 23(2), 2005Vol.23and bottom is stronger, resulting in bigger differenceof bottom decreases, less noise intensity differencebetween the soft bottom case and hard bottom case.is observed between the far field intensities for softAs shown in Fig.7b and c, for winter SSP, the effectbottom and hard bottom case.Constant SSPDownward refracting SSP0*0°hard30"30°soft。soft60250%0"50f40[30/2050(40/30 ,po°L-90° 9o"20~J -90Level (dB)Upward refracting SSP-30°soft!s0°J -90"Fig7 The vertical beam distribution of wind generated noise with soft or hard bottom at 500 Hza. for constant SSP; b. for summer SSP; c. for winter SSP30*constantfP不ummer-60°s090°40302010.90° 935 2020 3390°Fig.8 The noise vertical beam distribution under the constant SSP versus summer SSP at 500 Hz and soft bottoma. our model; b. Yang's model (constant SSP- solid line, summer SS--dotted SSP)b中国煤化工90*3520-135-9° 90YHCNMH G23590°Fig.9 The noise vertical beam distribution predicted by Yang's model for winter SSPa. soft bottom; b. hard bottomNo.2LIN et al: Model of wind-generated ambient noise in sratified shallow water151_0°ba0"30"30°hards0%人60*600°f50(40/302o710/640209(.90*Level (dB)Fig.10 The noise vertical beam distribution predicted by our model for the winter SSPa. soft bottom; b. hard bottmSciences Press, Beiing.6 CONCLUSIONSFinette, s. and R. M. Heitmeyer, 1996. Angl-time frequencyresolution of the noise field generated by wind-inducedNormal mode method is appropriate for far fieldbreaking waves. J. Acoust. Soc. Am 99(1): 209-222.calculation, while ray method is appropriate for nearGao, T. F. and E. C. Shang, 1982. The transformation betweenfield calculation. Combining the advantages of the twothe mode representation and the generalized ray repre-method, we suggest a hybrid wind- generated noisesentation of a sound field. Journal of Sound and Vibration80(1): 105-115.model of normal mode method (for far field) and rayHamson, R. M, 1997. The modelling of ambient noise due tomethod (for near feld) in stratified ocean. Consideringshpping and wind sources in complex environments. Apthat the near field of wind- generated noise is not sen-plied Acoustics 51: 251-287.sitive to the change in SSP, we replace the SSP withHamson, R. M, 1985. Vertical aray response to shipping noise:isovelocity SSP, making the new model runningMode/measurement comparisons for range - dependentrapider with some accuracy. We predict the effects ofMediteranean sites. J. Acoust. Soc. Am.94: 386395.Hamson, R. M, 1985. The theoretical responses of vertical andsuch parameters as fequency, SSP, bottom feature onhorizontal line arays to wind-induced noise in shallowambient noise intensity and vertical directionality; thewater. J. Acoust. Soc. Am.78: 1 702-1 712.simulated results using our model agree quite well withHamson, R. M, 1984. The theoretical response of a vertical linethose of wave model.array to wind-generated noise in shallow water. SaclantCentre Report SR-82.Hamson, R M. and R A. Wagstalf, 1983. An ambient noise7 ACKNOWLEDGMENTSmodel that includes coherent hydrophone summation forsonar system performance in shallow water. Saclant CentreThe authors are grateful to the crew of theReport SR-70.Underwater Acoustics Research Center, Institute ofHarison, C. H, 1997. CANARY: A Simple Model of AmbientAcoustics, Chinese Academy of Sciences and LiNoise and Coherence. Applied Acoustis 51: 289-315.Xuejun from Qingdao Acoustics Lab of Institute ofBreeding, Er, 1994. RANDI 3.1 User's Guide. AD-A286 300.Acoustics, Chinese Academy of Sciences for theirKuperman, W. A. and F. Ingenito, 1980. Spatial correlation ofsurface generated noise in a stratified ocean. J. Acoust. Soc.support and help.Am. 67: 1 988-1 996.Ligett, W. S. and M. J. Jacobsen, 1965. Covariance of surfaceReferencesgenerated noise in a deep channel. J. Acoust. Soc. Am.38:303-312.Talham, R. J, 1964. Ambient-sea noise models. J. Acoust. Soc.Buckingham, M. J, 1980. A theoretical model of ambient noiseAm.36: 1541-1 544.in a low-loss shallow water channel. J. Acoust Soc. Am.67:Yang, T. C. and Y. Kwang, 1997. Modeling the environmental1186-1192.influence on the vertical directionality of ambient noise inCron, B. F. and C. H, Sherman, 1962. Spatial crrelation func-中国煤化工America 101(5):tion for various noise models. J. Acoust. Soc. Am. 34:1732-1736.CNMHGWang, D. Z. and E. C. Shang, 1980. Underwater Acoustics.